Performance of Multiple True or False Options with Negative Scoring Schemes Against Single Best of Five Answer Options in a Medical School.

BACKGROUND AND OBJECTIVES The multiple true or false (MTF) options with different negative scoring schemes and the Single Best of Five Answer (SBOFA) option of multiple choice questions are used in assessing knowledge in medical schools in Nigeria. There are diverse reasons for specific institutions to use a given scheme. These include large candidates' population, time pressure on those conducting the examination and limited resources. This study was set out to compare the performance of two negative scoring schemes of the MTF pattern of objective questions with the SBOFA pattern as the standard objective questions pattern in a medical school Methodology: In this cross-sectional comparative study, selected response items were administered to medical students. A set comprised of SBOFA options (taken here as A-score) and the other two, MTF options with different negative marking schemes; negative marking (-1/2) as B-score, and negative marking (-1) as C-score. The mean scores were compared among the three schemes. Cutoff or pass mark was 50% across-board. Correlation statistics, Bland Altman plots and ROC analyses were used to compare the two negative scoring schemes of the MTF pattern with the SBOFA pattern as the standard objective questions pattern Results: The mean scores of the students for the schemes were A-score 42.9%, B-score 45.6% and C-score 35.0%. The correlation between A-score and B-score was significant (r=0.351, p=0.009). A-score also correlated significantly with C-score, (r=0.381, p=0.004). In B-score versus A-score the bias was -2.6 (47.4%), limits of agreement -29.4 - 24.1 (20.6% -74.1%). In C-score versus A-score the bias was 8.1 (58.1%) and limits of agreement -21.5 - 37.7 (28.5%-087.7%). For B-score and A-score area under receiver operator curve (AUROC) was 0.720 and for C-score and A-score 0.714. B-score at best cutoff mark 45.0% (sensitivity 93%), predicted A-score 50.0%. C-sore at best cutoff mark 33.0% (sensitivity 93%) predicted A-score 50%. CONCLUSION The MTF objective questions with negative marking scheme of -1/2 as penalty and the SBOFA scheme were more closely related than the MTF with -1 as penalty and SBOFA option. The SBOFA option had better correlation with the MTF with -1 as penalty while the -1/2 penalty had a much higher pass rate.